Palmeri Teaching & Projects @ Duke BME
This website contains syllabi and content for the courses that I teach in medical device design, including summaries of past and current student projects. I will also highlight recent research in acoustic radiation force imaging, tissue characterization with shear waves, and deep learning in medical imaging, and occasionally I’ll upload some blog posts.
Teaching
I actively teach Medical Device Design I & II (BME473L & BME474L) as part of the BME Design Fellows program and senior capstone design. You can find a summary of some student design projects in the Design Portfoilo.
I have also taught several other courses in the department, including Introduction to Medical Instrumentation (BME354L), and I have developed courses, including Medical Software Design (BME547) and Biomedical Circuit Analysis (BME790L).
Latest Project News
New preprint paper on a deep learning approach to help screen children for preventable hearing loss: A Hybrid Deep Learning Approach to Identify Preventable Childhood Hearing Loss”
Felix has a new preprint and associate software repository about SweiNet, his deep learning model to estimate shear wave speed in cervix to screen for premature ripening: “SweiNet: Deep Learning Based Uncertainty Quantification for Ultrasound Shear Wave Elasticity Imaging” GitHub: https://github.com/fqjin/swei-net
Felix has a new paper detailing his use of deep learning to guide skin segmentation: “Semi-automated weak annotation for deep neural network skin thickness measurement”. The associated software and Jupyter notebook can be found here.
Check out the latest publication from the RSNA QIBA US SWS effort studying the bias between different manufacturer shear wave speed systems in elastic and viscoelastic, tissue-mimicking phantoms: “Radiological Society of North America/Quantitative Imaging Biomarker Alliance Shear Wave Speed Bias Quantification in Elastic and Viscoelastic Phantoms”.
A new publication on using shear wave dispersion in to characterize cervical remodeling: “Shear Wave Dispersion as a Potential Biomarker for Cervical Remodeling During Pregnancy: Evidence From a Non-Human Primate Model”.
Open-Source Software and Data Repositories
We highly value generating easily-accessible code and data for the research community. This list highlights some of our latest shared contributions:
Acoustic Radiation Force FEM Tools: Templates for creating simulations of ARFI/SWEI excitations in elastic, viscoelastic and anisotropic media. LS-DYNA is the primary solver used in these simulations.
Ultrasonic Tracking Simulation Tools: Field II-based tools to simulate ultrasonically-tracking ARF-induced motion in ARFI/SWEI.
Verasonics Phantom Sequences: Sequences and post-processing code to evaluate shear wave speed group and phase velocities in elastic and viscoelastic media (part of the RSNA QIBA US SWS effort). Verasonic Processing Test Data can be found in the Duke Digital Repository.
MimickNet: Matching clinical-grade ultrasound post-processing without the hassle. Partially-beamformed and processing ultrasound data is available through the Duke Digital Repository. This work is led by Ouwen Huang.
Dermis Segmentation: Two methods for segmenting skin on ultrasound B-mode images. This work is led by Felix Jin.
QIBA Ultrasound Shear Wave Speed Digital Phantoms are available online for validating SWEI algorithms in elastic and viscoelastic media. The code associated with generating these data can be found here: https://github.com/RSNA-QIBA-US-SWS/QIBA-DigitalPhantoms.
A comprehensive list of all of our online shared resources can be found here: https://mlp6.github.io/projects/open-source-software-data.
COVID-19 Engineering Response Team
I have been involved with a highly-collaborative, inter-departmental effort within the Pratt School of Engineering to help our medical center and campus deal with the COVID-19 pandemic. You can read more about our efforts here: https://pratt.duke.edu/covid19/response-team.